How Ford Valencia used Twinn Witness to shape AGV strategy

Ford Valencia initially identified the potential of AGVs to optimise the Kuga door assembly process. Previously, transportation from closure sub-assemblies to the body assembly line had been carried out manually by forklift operators. And it was at this early stage that Vicent realised the world of AGVs was full of contradictions.

"The first thing that surprised me was that each AGV manufacturer who came to bid on the project proposed a different number of AGVs," he recalled. "One said we needed 6, another claimed it could be done with 3, while another still said it was impossible with fewer than 12."

Too many AGVs, and the project cost would be inflated; too few, and production would be endangered. To ensure Ford Valencia invested in the correct number, they needed to account for many dynamic variables – blockages, unexpected breakdowns and stock control, for example. After results using traditional methods proved unreliable, he recognised the need for simulation.  

"I’m not a simulation specialist, but I was aware of the software. Ford Cologne uses it to simulate manufacturing flows in new models, for example," explained Vicent. "With our first model simulation, we determined that the minimum viable quantity was 10 AGVs – 5 to transport the front right doors and 5 for the left," he continued. "And we have been working with that approach without any cadence problems."

This was a team effort that included leadership from Javier Leiva, Senior Industrial Engineer from the Productivity Department: "He gave the green light to the project based on that initial model, despite the doubts everyone had ¬about the new approach. Without that trust, we wouldn’t be where we are today."

Ford soon encountered another paradox, which they could unpack using the power of Witness simulation. To meet targets, Ford needed to increase its Kuga production by 10% per shift. "To achieve that, the logical conclusion is that you’d need to increase the number of AGVs," explained Vicent. “However, it actually depends on the circuit – our front door transportation circuit is a carousel, with a point of origin and a destination.”

Using Witness’s Experimenter functionality, they discovered that increasing the number of AGVs increased the number of AGVs waiting in line. "Adding more AGVs to the circuit actually showed an increase in production time of 3 seconds."

"Witness allows you to choose the best scenario, setting the goal and playing with the variables including the number of AGVs – we ran it with 4, 5, 6, 7 and 10 – and the production volume was the same."

Having saved money by not buying unnecessary AGVs, Ford was instead able to invest in decreasing cycle times and improving the availability of lines to meet its production target.

However, using Witness, they demonstrated that both AGVs were required. By simulating the same circuit with just one AGV, they proved that the production cadence would not be met, resulting in a 10% line stoppage.

The conclusion? "With AGVs, what seems logical is not always the case – sometimes the AGV cannot be reassigned to another mission," Vicent explained. 

This example also illustrates another key benefit of simulation: "Knowing where AGVs will be idle for longer periods is useful for determining where to install battery chargers, which data manufacturers don’t usually have until the project is underway. This reduces installation times, which in turn delivers cost savings."

What’s more, Witness has also enabled Ford to quickly calculate AGV numbers required for large-scale projects. "By creating simple templates, we can estimate how many AGVs we need for new vehicle launches, for example – generating different possible scenarios quickly to empower evidence-based decisions," he added.

After years of studying AGV circuit behaviour – with Witness’s help – Ford has developed a nuanced understanding of this complex world: "An AGV circuit is not like a conveyor, a carrier system or a robot station. It interacts with its environment, has its own characteristics and requires a new way of working, to which we were not accustomed."

He added: "Production doesn’t take place in laboratory conditions. Using Witness, we’ve demonstrated that – while we could meet transport requirements for a certain task with just 4 AGVs, for example – we’re better off in the longer term investing in 5 that work at a lower capacity."

Vicent still recalls the first time he used Witness to simulate a body plant project: "For me, it was like having a crystal ball. We could see in advance how door stock was drastically reduced, yet we could still service the line and maintain production cadence. Before the AGVs, we stacked the door racks to have stock in case of breakdown. The simulation showed us we could reduce the quantity by 50%."

Some simulations have been relatively simple, while others have involved modelling complexity. The third AGV circuit required Ford to automate the transportation of the Kuga underbody, where AGVs had to deliver up to 18 different references.

To determine the most efficient combination, they needed support. And thanks to Helios Alvarez, Area Manager for Body & Stamping, there was access to training needed to create the complex simulation model.

This collaborative approach is the future of AGV and simulation at Ford: “Each plant at Ford Valencia shared our challenges with AGVs – and now we have to adopt the same approach to enable swift progression across all European and global plants."

Progress is accelerating thanks to the support of Paula Carsi, Transformation and Innovation Manager. "Paula believes in the power of simulation to achieve accurate results. The innovation department is growing exponentially, with all 4 Valencia Site Plants benefitting from knowledge sharing and site visits. Work on digitalisation projects – including digital twins and shadow twins – is also underway."

They are now creating in-house courses so others can build simulation expertise:

• In 2021, the AGV Innovation team created a training course on using Witness for AGV
• In 2024, the team is creating introductory simulation training courses for engineers from all plants – and for any industrial process

In addition, Ford Valencia has requested help from Twinn to create a small digital twin for one line in the body shop. The goal? Short-term results that answer the question: "With current line conditions, will I achieve what I want to by the end of the day?"

More about Ford Valencia

The Ford Valencia Body and Assembly Plant is Ford's largest manufacturing plant outside of the United States of America and has a production capacity of 450,000 vehicles per year.

Recently, the team received a Ford Alliance Manufacturing Innovation Team 2023 Award in the Category of Highest Savings Project. This project demonstrated the automation of transporting Ford Kuga underbody parts using AGVs.

To read more about the use history and use of simulation at Ford, please click here.